Interior Designers: YOUR AI Results Are Mediocre (And It's Not the Prompt)
The three layers that determine AI output quality
Interior designers using AI get generic results for one reason: AI knows nothing about your business, your clients, or your design voice unless you tell it first. Prompt engineering, the skill every tutorial focuses on is the least important of three layers that determine AI output quality.
Intent (the outcome you want) and context (everything AI needs to know about your situation) come first. Get those two right, and the prompt almost writes itself. Get them wrong, and better prompting just polishes something that was never going to work.
Apple - Youtube Music - Spotify - Youtube
-
[00:23.1]
If you've been using AI and getting results that are fine, but nowhere near what the big AI companies are telling us we should be getting from AI I'm here to tell you the problem isn't the AI. The problem is us.
[00:41.2]
I want to introduce you to somebody. I'm going to call her Maya. She's an avatar of a group of, interior designers that I've been working with over the past few months. Maya's been in the business for, say, two or three years.
[00:57.5]
She's either a millennial, or possibly even a Gen Z. She's very comfortable with technology. She is excited to work with AI and see how that can improve her career working within an interior design firm or in her own interior design firm that she has started as a solopreneur.
[01:19.0]
She watches AI tutorials on YouTube. She reads all sorts of stuff about AI. She's taken courses, she's bought products. She's doing everything that she is supposed to do. And like most of us, her results are okay.
[01:37.5]
They're not bad. She's not quitting AI but they're not actually good enough for her to trust AI fully. Especially with things that are client facing in her business or in her the business she works for, she uses it for social media captions, she uses it for emails, she uses it for research, she uses it for image generation, for concepts.
[02:07.0]
There's probably some work with rendering. But, anything. Client facing, anything that really represents her business. No, it's not happening. Because what she's getting from AI doesn't sound like her, doesn't look like the design she would actually produce.
[02:29.2]
So Maya, does what a lot of interior designers are already doing. She assumes AI just isn't ready for prime time. It's not ready to play a significant role in her business or the business in which she works. Or she assumes that maybe she's the problem.
[02:45.1]
She's not technically advanced enough to get out of A.I. what? She wants to get out of A.I. and blames herself. And here's the thing. Maya's 100% wrong about that. And I say that with, the utmost respect.
[03:01.8]
Because over the past nine months, I have had over 100 conversations with interior designers in regards to incorporating AI into their business and how they could or should be working with AI. The AI The AI Is amazing.
[03:19.5]
The tools are capable of doing fantastic things. The problem is what Maya, is bringing to the conversation between her and her AI Whether that's Gemini or Claude or Chatgpt or whatever. More accurately, it's what Maya is not bringing to the conversation.
[03:40.8]
And what do I mean by that? There are three things that determine your AI output quality. And I want to be very clear. Those three things are not equal. There's a definite hierarchy to those three things. And most interior designers, like our avatar Maya, are investing almost all of their energy in only one of those things.
[04:05.7]
And that thing is prompt engineering. This is the one we've all heard about. It's the words you use when you ask AI something. How you phrase the question, how specific you are, what format you ask for. I like to think of it like how I learned how to use Google Search way back in the day.
[04:27.8]
Back in 1999, everyone typed full sentences into Google and got kind of crappy results over time. With practice, we all learned shorter, more specific, better terms, things that made Google work better for us.
[04:45.5]
And AI has a similar learning curve, except it's actually way more forgiving than a, Google search, because natural language actually works. You can talk to it like a person, but when we learn how to talk to it in a specific way, it.
[05:02.6]
It does still improve the results we're going to get. And here's the uncomfortable part. Prompt engineering is where most people spend all of their time and energy. And it's the least important of the three layers that I want to discuss. If the other two layers are weak, better wording, better prompting doesn't save you.
[05:23.7]
You're essentially polishing the surface of something that doesn't have a solid foundation. The second layer I want to discuss is context. And this idea of context engineering is starting to creep into the mainstream. It's still mostly an AI nerd topic, but it is becoming discussed.
[05:42.4]
If you search YouTube, you will find things on context. Engineering context is everything that you tell AI about your situation before you ask your actual question. And this is where most interior designers are leaving enormous value on the table.
[05:57.6]
Because I knows absolutely nothing about you unless you tell it. Nothing about your business, nothing about your clients, nothing about your market, your style, your price point, your voice. Every conversation starts from zero. Think about it this way.
[06:13.4]
Imagine you hired a brilliant new junior interior designer, just graduated from design school, incredibly talented, genuinely excited to work for you. Like you see yourself in them, right? It's going to be amazing. Their first day at work, there's no way you're going to hand them a project brief and walk away and let them get to work on it.
[06:34.0]
You'd sit down with them, you'd explain who your clients are, what matters to them, how you work, what your aesthetic is, what you've tried that didn't work, what you tried that did work. Your past history, as when you were a young designer, you'd give them the background that they need to do good work for your business.
[06:53.1]
And that process of teaching them to design in your style would take weeks and months and possibly years. AI is exactly the same. Brilliant, capable, and completely unaware of your world until you fill it in with that information.
[07:12.0]
When you give AI real context, your business situation, your client profile, your constraints, your voice, your strengths, your weaknesses, it has way more raw material to work with. And as a result, it can give you something relevant instead of something generic. That's not magic, that's just information.
[07:29.9]
The more specific and honest the information, the better output you're going to get from any AI. The third layer I want to discuss is intent, and this is the one that almost nobody is talking about yet, which is exactly why I want to spend some time on it here.
[07:47.8]
Intent is not the task. Intent is the outcome we want. There's a big difference. Write a social post to get me new clients is a task. I need an Instagram post that will make a homeowner in my geographic area stop scrolling because they recognize their own frustration and might just see me as the solution to that frustration.
[08:08.7]
That's an outcome. That's intent. Most people skip this step entirely and go straight to just asking a question of AI. And that's where the output falls apart. So here's how you can think about these three things.
[08:24.9]
Intent is the foundation. It's what you're actually trying to accomplish. Context is the enabler. It's everything AI needs to know about your situation to give you a relevant answer. And prompt is the execution. It's how you phrase the actual ask of AI.
[08:43.4]
Get intent and context. Write and the prompt almost writes itself. Focus on the prompt while intent and context are weak and you're working very, very hard to get mediocre results. Let me give you another example. Say you want to reach out to potential new clients.
[09:02.0]
Here's what most people do with AI. You open up ChatGPT or Claude or whoever and type help me write an outreach email to potential clients and you get something back that's technically correct and completely useless for actually getting new clients.
[09:20.1]
It's professional, it's generic. It sounds like it was written for nobody in particular because it was. Here's a better way to do it. Before you type anything, you answer three questions. What am I actually trying to accomplish here?
[09:35.8]
Not just send an email, but what does success look like? Who specifically am I trying to reach, and what do I know about them? And what does good look like? What tone? What length? What feeling do I want to leave them with? Then you go to AI and you say something like, I want to reconnect with people who've seen my work but haven't hired me, yet.
[09:59.2]
I work with busy professionals in New York City renovating their primary residences. They're usually in their 40s or 50s with no kids. They care about quality. They're not chasing trends. They value privacy and are willing to pay. My tone is warm but direct.
[10:15.7]
Write me a short message under 150 words that. That opens a door without feeling pushy. Same tool, same AI. You will get a completely different output. The difference was entirely in what the human brought to the conversation, the input from the human being.
[10:35.4]
But here's the reason why I'm creating this video. When I explained this concept of intent and context engineering. When I explain this to my interior designer clients, and I've had this conversation a lot over the past year, most of them get it.
[10:52.2]
They understand the theory, they nod, and then I get them to try it. And a lot of them still don't get the results they're looking for. Not because the framework is wrong, but because of something that the big AI companies and AI experts and YouTube influencers are not addressing.
[11:12.0]
To explain this problem more completely, I want to introduce you to another client, avatar Carol. Carol is a, More experienced designer, 22 years in the business, built her firm entirely off of referrals. She has a reputation, she has a process.
[11:27.7]
She has a way of working with clients that's deeply her own. She's played with AI over the last two years, has got results that feel generic and perhaps even slightly insulting to her expertise, and more or less set it aside.
[11:44.6]
She came to this conversation not because she believed in AI, but because 2025 was a down year for a lot of interior designers. Same for Carol, and she was willing to try things that she wouldn't have considered otherwise. So Carol asked me to come in, and I sat down with her and I explained everything I just explained to you, and she got it.
[12:07.5]
She understands the framework, intent, context, prompting. She knows exactly what she wants from working with her AI, but. When it comes to context, things got kind of Messy. Carol has 22 years of context.
[12:25.4]
She knows exactly who her best clients are. She knows what makes a project work and what makes it falls. Fall apart. She knows what she's worth and why she's worth it. But when she tries to put that into words for her AI, she hits a wall. Not because she doesn't know it.
[12:43.0]
After 22 years, that knowledge is so deep inside her that she's almost stopped being able to see it from the outside. It's professional intuition at this point. Pattern recognition. She's Michael Jordan shooting basketballs.
[13:00.2]
It's never been written down, or even if it has been written down in some sort of systems bible, it does not exist at the level that it exists in her head. And this is not a unique problem.
[13:16.0]
It's the exact problem that I ran into when I started really putting time into understanding and working with AI for my business. And it's why I started building Socratic AI tools for my business and my personal life.
[13:33.2]
AI tools that stop AI from immediately offering solutions and force it to ask questions. Questions. Questions that draw the context out of me, out of you, so that we can share it with the AI and give it the ammunition it needs to offer way better solutions.
[13:52.7]
So let's go back to our avatar, Carol. When I got Carol to stop what she was doing and got her to use the tool that I built for her business, everything completely changed with how she worked and used and the results she got from her AI.
[14:08.5]
The tools forced her AI to ask her questions she would never have thought to answer on her own. They asked about her best clients, not just who they were, but what made those relationships work, the psychographics, all that kind of stuff. They asked about decisions that she'd made on projects that didn't go the way she'd planned.
[14:28.5]
They asked what she knew now that she wished she'd known 10 years ago, 15 years ago, 20 years ago. Questions that drew out what Carol knew but had never properly articulated in her own systems and processes in her business.
[14:46.7]
And at the end of the session, Carol looked at what had come out of that conversation with her AI and said something that I've heard variations of more than once. She told me that I've been trying to explain my value to clients for years, and I couldn't. But here it is, 22 years of expertise, finally in a form that AI could actually work with, finally in a form that she could use in her business.
[15:13.9]
And every time she uses this tool that I built for her, her AI remembers her AI learns more and more about Carol and the answers keep getting better and better and better. Like I can only imagine after a year or two of using this and the improvements that AI is going to see what she is going to be to do, where she'll be actually able to let AI go and do things for her.
[15:40.2]
And I, I need to be clear about what happened here. The AI didn't get smarter. Carol didn't suddenly become more articulate. What changed is that she finally had a way to get what she knew out of her head and into the conversation with her AI.
[15:56.7]
That's the gap the framework alone doesn't close. The framework of intent, context and prompt engineering Understanding prompt, context and intent is necessary 100%. And if that's all you take from this video, that's fantastic.
[16:12.5]
You are ahead of almost everybody out there. But if your best professional expertise has never been put into words anywhere, and for the most experienced designers it probably hasn't, you can't manually transfer has to be drawn out of you.
[16:28.7]
It's like going and visiting a therapist and they pull things out of that you completely had forgotten about. Carol's story is a perfect example of why AI isn't really working for interior designers. And not just interior designers for all business owners. Not just because interior designers don't truly understand AI.
[16:50.1]
Because the most valuable thing a designer brings to any conversation with an AI, their expertise, their judgment, their professional instincts, built over years and years of working with clients and making mistakes and getting better, almost never makes it into the conversation.
[17:08.4]
Not because designers are doing something wrong, because nobody ever needed to put that knowledge into words before. You just knew it. And now if you're going to work with AI, it needs you to tell it that stuff to get better results. The tools I built, my Socratic tools ask questions to draw out, what you know, your client instincts, your positioning, your professional judgment in a way that AI can actually use it.
[17:35.7]
It's not a shortcut, it's a structured way to do the thinking that determines whether AI gives you something genuinely useful or something that sounds like it came from a business school textbook or some influencer off of YouTube.
[17:53.4]
So whether you're like the avatar Maya or the avatar Carol or somewhere in between, as an interior designer, remember that prompt engineering isn't the solution to getting the most from AI. AI needs to really, really, really understand your intent and context to give you something special.
[18:14.9]
How, you provide that intent and context is up to you. In the coming weeks, I'm going to produce a bunch more videos specifically on intent engineering and context engineering for all of you that want to DIY your own AI in your business and in your personal life, like I did with Carol, I also work one on one with interior designers to help them integrate AI into the business side of their business.
[18:39.3]
So feel free, reach out, have a discussion. I also sell a variety of AI tools that these Socratic AI tools designed specifically for interior designers. These are the same tools that I used with Carol in our one on one.
[18:54.9]
You just won't have me by your side as your AI Sherpa like Carol did. So I'm hoping that this video is going to help you understand context engineering and intent engineering and help you get more out of your AI.
[19:12.1]
Like I said, more videos to come. I hope this helped.
Key Takeaways
There are three layers that determine AI output quality: intent, context, and prompt. They are not equal — intent is the foundation, context is the enabler, prompting is the execution.
Prompt engineering is where most designers invest all their energy. It's the least important layer. Better prompting cannot save weak intent and context.
AI starts every conversation from zero. It knows nothing about your clients, your style, your price point, or your voice — until you provide that information.
Intent is not the task — it's the outcome. "Write me a social post" is a task. "Make a homeowner recognize their frustration and see me as the solution" is intent.
The most experienced designers often struggle most with AI. After 20+ years, professional expertise becomes intuition — it has never been written down, and it can't be manually transferred. It has to be drawn out.
Socratic AI tools — designed to ask questions before offering solutions — draw out the context that makes AI genuinely useful versus generically competent.
Conclusions
Strategic Benefits for Your Interior Design Business
Interior designers who build strong intent and context foundations get AI outputs that actually sound like them — client-facing copy that represents their aesthetic, their voice, and their market positioning. The difference between AI that saves hours per week and AI that creates more rework than it prevents comes down entirely to what you bring to the conversation before you type anything.
Implementation Blueprint
Before your next AI session, answer three questions in writing: What am I actually trying to accomplish, not the task, the outcome? Who am I trying to reach, and what do I specifically know about them? What does good look like — tone, length, the feeling I want to leave them with? Then give AI your business context: your client profile, your market, your price point, your voice, what you've tried that didn't work. The more specific and honest the input, the more relevant the output.
Professional Transformation
The goal isn't AI that does more work, it's AI that does your work. Work that represents your taste, your clients, your business. That shift happens when AI finally understands the context that lives inside your head. Not through better prompting. Through better preparation.
Join the Interior DesignHer Community
If you enjoyed this guide and podcast, you're going to love all the other business education resources we're bringing to interior designers.
As interior designers ourselves, we created Interior DesignHer to bring top-notch business education to interior designers. We especially have a soft-spot for interior design solopreneurs & micro-businesses.
Subscribe now so you never miss our latest business guidance, guest interviews, and insider strategies specifically for interior designers.
